甘蓝型油菜硫苷组分近红外检测模型的建立

The Construction of a Near Infrared Inspection Model for Glucosinolate Constituent Contents in Brassica napus L.

为探索近红外光谱技术在快速检测硫苷组分质量摩尔浓度方面的可行性,该研究选取了甘蓝型油菜籽材料266份,采用高效液相色谱法测定了硫苷组分的质量摩尔浓度,并利用近红外分析仪和WinISIⅢ软件对数据进行判断与处理,建立了甘蓝型油菜籽硫苷组分的近红外检测模型.其中,2-羟基-3-丁烯基脱硫硫苷,3-丁烯基脱硫硫苷定标决定系数(RSQ)与交叉验证决定系数(1-VR)均超过了0.90(越接近1越理想),说明模型预测性能较好,可以用于油菜品种的快速筛选.但是,4-羟基-3-吲哚甲基脱硫硫苷,4-戊烯基脱硫硫苷,苯乙基脱硫硫苷模型的RSQ分别为0.470,0.219和0.147,1-VR分别为0.304,0.128,0.175,建模效果欠佳,还需要进一步研究.

To explore the feasibility of rapidly detecting the contents of glucosinolate constituents with near infrared reflectance spectroscopy, we selected 266 representative samples, and analysed their glucosino lates constituent contents with high performance liquid chromatography （HPLC） and processed the data, using an NIR spectrophotometer and the WinlSIl]I software. In the end, a near infrared inspection model was constructed for glucosinolates constituent contents in Brassica napus L. The results showed that the cross-validation correlation coefficient （1-VR） of the NIR spectroscopy model for desulfoprogoitrin （DSPRO） and desulfoglueonapin （DSGNA） was 0. 937 2 and 0. 900 9 and the external validation correla- tion coefficient （RSQ） was 0. 962 and 0. 937, respectively, which suggested that the model was satisfacto- ry and could be used for the rapid screening of rapeseed varieties. However, the RSQ of desulfo-4-hydrox- yglucobrassicin （ DS4-OHGBS ）, desulfoglucobrassicanapin （ DSGBN ） and desulfogluconasturtiin （DSGST） was 0. 147, 0. 219 and 0. 419 and their 1-VR was 0. 175, 0. 128 and 0. 304, respectively, which result was unsatisfactory and needed further research.